[PDFBOX-2847] mergeDocumentsNonSeq does not utilize scratchFile

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.8.9, 1.8.10, 2.0.0
• Fix Version/s: 1.8.10, 2.0.0
• Component/s: Utilities
• Assignee: Tilman Hausherr
• Reporter: Jim deVos
• Created: 30/Jun/15 04:32
• Updated: 23/Jul/15 06:35
• Resolved: 30/Jun/15 20:04

Description

I noticed when merging relatively large pdfs (1gb) that the heap would grow by at least the same amount until complete, even when I call mergeDocumentsNonSeq() and supplying a read/write scratchfile.

When I looked at the source for mergeDocuments(bool, RandomAccess), it looks like the scratch file is never used.

private void mergeDocuments(boolean isNonSeq, RandomAccess scratchFile) 
            throws IOException, COSVisitorException
{
//...snip

                if (isNonSeq)
                {
                    source = PDDocument.loadNonSeq(sourceFile, null);
                }
//...snip
}